[tmql-wg] TMQL comments
rho at bigpond.net.au
Wed Jan 10 00:32:39 EST 2007
On Fri, Dec 29, 2006 at 05:09:42PM +0100, Lars Heuer wrote:
> 2 Normative references
> The reference should be: "13250-6"
> Additionally, the official name is "Topic Maps - Compact Syntax"
> 3.2 Ontological Comittments
> Should be: 'tmdm' or (better) 'dm'?
At the moment I have here tmdm. 'dm' reminded me too much of
'Drogeriemarkt', a drug store chain in Austria. But I'm not
religious about that ;-)
> XQuery and some other XML-related sources refer to the XML Schema
> Datatypes with the 'xs' prefix. I am not sure if we should also adapt
> the 'xs'
Will have a look, ok.
> 4.2 Atoms
>  boolean -> True | False
> Should be: true | false since 'True' and 'False' are not
> valid values for xs:boolean.
> (From a programming language designer POV I understand why the terms
> are capitalized (they are constants)), but I think the alignment to
> xs:boolean is more important).
I have already changed them to 'true' | 'false'.
> 4.3 Item References
> If the item reference is an identifier (without prefix) then this
> identifier is interpreted as a item identifier for a topic in the
> context map. The result is then this topic item. If no such topic
> exists, an error will be flagged.
> Why is such an reference 'just' interpreted as item identifier? IMO it
> would be better, if such an reference is interpreted as 'Either a topic
> with the specified subject identifier OR item identifier', since there
> can only be one topic with either the item identifer or subject
Just to be clear on this. We have the following cases:
rumsti .... this is interpreted as item identifier
http://rumsti.com .... this is interpreted as subject identifier (indication)
http://rumsti.com ~ .... the same, just more explicit
some:rumsti .... also subject identifier
some:rumsti ~ .... the same, more explicit
http://rumsti.com = .... subject locator
some:rumsti = ... still subject locator
Yes, the text can be more streamlined to say exactly this.
> 2nd example:
> The following content is interpreted first as a URI value. The
> trailing = then signals that the URI should be interpreted as source
> locator for a topic in the context map.
> I think that is a typo. You mean "subject locator".
Definitely 'subject locator'.
More information about the tmql-wg